"Grow Old Along with Me":
>The erosion of marriage is a constant refrain in political debate and
a legitimate concern for society in general. Recent research by Lee Lillard
and Linda Waite indicates that it also has severe consequences for individuals.
Both men and women benefit from being married and both are at greater risk
of dying, at any age, when they are not married--whether never married,
separated, divorced, or widowed. In "`Til Death Do Us Part': Marital Disruption
and Mortality," Lillard and Waite describe a study in which they provide
new insights into the "marriage benefit" and how it works.
Reaping the Marriage Benefit
This study uses a sample of more than 11,000 individuals from the Panel
Study of Income Dynamics to focus on the "protective effect" of marriage.
It examines how risk of dying can change with changes in marital status
and over time and how much of the marriage benefit is explained by three
key factors: financial well-being, living arrangements, and marital history.
"For Richer or for Poorer"
Most previous studies of marriage and mortality have found that marriage
benefits men more than women--at least to the extent that the difference
between death rates for the married and unmarried is substantially higher
for men than for women. However, in this research, Lillard and Waite found
that, over time, marriage had about the same effect on mortality risk for
men and for women--but "over time" is the important qualification. Over
a long marriage, the effects for men and women even out, but at the beginning
there are significant differences. Right after the wedding, the risk of
dying drops considerably for men, but not for women (relative to the risk
for unmarried men and women). In contrast, the growing benefit over time
is somewhat larger for women than it is for men.
The authors speculate that this immediate drop for men might be
explained by premarital "lifestyle": Unmarried men are more likely than
unmarried women to engage in risky behavior (for example, poor diet, immoderate
drinking, belligerent social behavior). Marriage often brings a more settled
lifestyle and more moderate behavior--not to mention better nutrition.
But what accounts for the greater cumulative benefit for women over many
years of marriage?
The study's results suggest that improved financial resources
are a key avenue through which marriage improves well-being and life chances
for both men and women--but the effect is much greater for women than for
men. Men's risk of dying decreases significantly with marriage--even at
low income levels--but married women's risk does not drop significantly
until income reaches higher levels. The authors conclude that if "women
benefit to an important extent from the access to higher household income
that marriage gives them, then this income may be buying them access to
better health care, better nutrition, better housing, a safer job, less
physical and mental stress, and so on."
Other effects on longevity are less clear from the study's results.
Married men's chances of dying drop as their wives' level of education
rises. This may indicate that better-educated wives run more health-protective
households. But it may also indicate that better-educated women tend to
select healthier and more stable mates. At any rate, living arrangements,
other than the married state, have no significant effect on longevity for
men or for women, once other factors are held constant.
Reverting to the "Single" State
When marriage ends, the effects demonstrate, again, how potent the marriage
benefit is and the central role of income for women. The figure's dual
panels provide an interesting contrast. Panel A shows the relative risk
of dying for men, by marital status. Men who are widowed, divorced, separated,
or never married face about the same risk of dying--and it is much higher
than the risk for married men. As Panel B shows, the analogous risks for
women differ in provocative ways. Currently married women face a lower
risk of dying than those who are divorced or were never married and a very
much lower risk than women who are separated. Most interesting is that,
unlike widowed men, widowed women have about the same risk of dying as
currently married women.
These results for widowed women may offer further evidence about income,
marital status, and mortality. Lillard and Waite sum it up as follows:
If, as our results suggest, marriage improves the life chances
of women primarily through improving their financial well-being, why do
never-married, separated, and divorced women fare substantially worse than
widowed women, once we take household income into account? We speculate
that widowed women with the same level of household income as divorced
women are actually better off financially, since they more often have access
to assets that remain from their marriage, especially a house. Divorced
women often lose their house and generally must divide assets with their
exhusband[s]. . . . [T]he superior asset position of widows compared to
divorced women or never-married women could account for the mortality differentials
we observe. The impact of assets, as well as income, on the mortality of
previously married women deserves additional attention.
When marriage ends, only widowed women seem to retain
of the protective effects of marriage
RAND research briefs summarize research that has been more fully documented
elsewhere. This research brief describes work done in RAND's Labor
and Population Program as part of the Center for the Study of Aging,
supported by a grant from the National Institute on Aging. The work is
documented in "'Til Death Do Us Part: Marital Disruption and Mortality,"
by Lee A. Lillard and Linda J. Waite, RP-487,1996
(reprinted from American Journal of Sociology, Vol. 100, No. 5,
March 1995, pp. 1131-1156). This document is available from RAND Distribution
Services (Telephone: 310-451-7002; Fax: 310-451-6915; or Internet: firstname.lastname@example.org).
Abstracts of RAND documents may be viewed on the World Wide Web (http://www.rand.org).
RAND is a nonprofit institution that helps improve public policy through
research and analysis; its publications do not necessarily reflect the
opinions or policies of its research sponsors.